Major advances have been made in the current generation of HIV preventive interventions. Theory-based interventions focused on individuals have demonstrated efficacy in reducing HIV/STD risk behaviors. However, there is growing concern that individual- oriented approaches are insufficient for modifying behavioral norms and affecting disease transmission at the community level, especially among those most at risk and least likely to volunteer for individually-based interventions. While behavior theories and recent empirical studies suggest that outreach interventions can be effective, few studies have evaluated the efficacy of social diffusion interventions, especially in the developing world. This application seeks to replicate the successful community public opinion leader HIV preventive intervention reported by Kelly et al. in community settings in Chennai (Madras), India. India is estimated to have more cases of HIV than any other country in the world, and is poised on the edge of a rapid escalation in their epidemic. Migrant workers (from the rural countryside) are numerous, especially in the southern capital of Tamil Nadu, Chennai, and have borne the brunt of the HIV epidemic to date, mostly associated with sexual behavior while separated from home. Migrants live in same-sex hostels in various industrial estates around the city, and can be targeted for a community public opinion leader intervention with a minimal concern of cross-contamination. With five years of support, we will accomplish the following specific aims: (1) To develop a culturally acceptable Community Public Opinion Leader (C-POL) HIV preventive intervention to reduce HIV/STD risk among young migrant workers in Chennai with our collaborators at YRG CARE in collaboration with the other AIDS Collaborating Teams and the Data Coordinating Center. (2) To collaboratively design and conduct a set of rapid ethnographic assessment procedures to determine the acceptability, content and format of the C-POL intervention in the Chennai venues. (3) To identify, recruit and train interviewer/assessors and conduct a survey of migrant workers from workplaces in Chennai with 100+ migrant workers who maintain single-sex hostels/residences on HIV/STD risk behaviors, peer influence, substance use, sexual history, and other theory-driven antecedent and proximal risk factors. Nine matched pairs of workplaces (based on industry, size and location) will be selected with 100 workers interviewed in each site along with donation of biological specimens (urine, self-administered swabs and fingerstick dot blots for gonorrhea, chlamydia, trichomonas and HIV infections). (4) To recruit, train and deploy C-POLs (n=125) for a period of one year in experimental venues; and (5) To evaluate intervention efficacy using a post-intervention sample of workers in the matched workplaces using a similar design as in Aim 3 (and then intervene in the comparison plants after the final assessment is complete), analyzing the intervention (main outcome, proportion of intercourse acts protected) using GEE approaches to longitudinal data analysis correcting for correlation within the data and adjusting for the design.